## Format

---

1. Title: Start with a concise and descriptive title that captures the essence of your research project.
2. **Introduction**: Provide a brief introduction to the topic of your research, highlighting its significance and relevance in the field.
3. **Objective**(s): Clearly state the objective(s) of your research project. What are you aiming to achieve or investigate?
4. **Methods**: Briefly describe the methods and techniques employed in your research. Mention the data collection process, any algorithms or frameworks used, and how you analyzed the data.
5. **Results**: Summarize the key findings and outcomes of your research. Highlight any significant insights, trends, or patterns that emerged from the analysis.
6. **Conclusion**: Present the main conclusions drawn from your research. Emphasize the implications and potential impact of your findings on the field of study.
7. **Significance and Contribution**: Discuss the significance and contribution of your research project. How does it advance the current knowledge or address existing gaps in the field?
8. Future Directions: Mention any potential future directions or areas for further research and development based on your findings.
9. **Keywords**: Provide a list of relevant keywords that represent the main concepts and focus of your research.

## Abstract

---

**Title: Learning Analytic Framework for Students' Academic Performance and Critical Learning Pathways Discovery**

Learning analytics has evolved as a prominent topic within education that uses data-driven methodologies to acquire insights into student learning and informed instructional practices. This study aims to develop a unified learning analytic framework which can be implemented by any institution to enhance the understanding of students' academic performance and critical learning pathways.

[Framework's features] The proposed framework offers distinct features that that captures the overall performance of students by considering multiple dimensions - grades, Grade Point Average (GPA), sequence of courses or subjects taken throughout the semesters or years of study. Its incorporation of a collective network graph 1) showcases the relationships and dependencies between courses, 2) highlights the emergence of specific pathways within an academic program, 3) analyse network centrality, 4) identifies critical nodes, 5) sheds light on influential courses that significantly impact overall student performance and learning outcomes.

set it apart from the existing approaches. One notable aspect is its ability to capture the overall performance of students by considering multiple dimensions such as grades, GPA, and the sequence of subjects taken by students throughout their study. Another unique feature of the framework is its incorporation of a collective network graph to showcase the relationships and dependencies between courses, highlighting the emergence of specific pathways within an academic program. By analyzing network centrality and identifying critical nodes, the framework sheds light on influential courses that significantly impact overall student performance and learning outcomes.

[Framework methodology] The proposed framework methodology encompasses several key steps which starts with a standardized data collection and data preprocessing process. Next, dimensionality reduction techniques such as Principal Component Analysis (PCA) and Non-negative matrix factorization (NMF) are applied to capture the most influential course components and grade information. The reduced dataset is then subjected to clustering algorithms, including Partition-based clustering (K-means), hierarchical clustering, and Density-based clustering (DBSCAN). These algorithms group students based on their academic performance and course profiles, allowing us to identify clusters of students with similar characteristics and academic trajectories. Furthermore, the critical learning pathways of the students are explored through the construction of a course network graph. This graph represents the sequence of courses taken by students and highlights the emergence of learning pathways within each program. Network analysis techniques are applied to identify influential courses and their impact on overall student performance.

In this case study, we focus on a sample of 3550 undergraduate students enrolled in three different programs at a private university in Malaysia. The dataset includes information such as student IDs, program of study, course names, total marks, grades, and session intakes. By leveraging the proposed learning analytic framework, we aim to gain valuable insights into the students' academic journeys and identify factors that contribute to their success.

Keywords: Learning analytic framework, unsupervised learning, Principal Component Analysis (PCA), Non-negative matrix factorization (NMF), educational data mining (EDM)